The AI-policy-governance nexus: How regulation and AI shift corporate governance toward stakeholders

Abstract

This study investigates how artificial intelligence (AI), when deployed under sustainability-oriented policy regimes (e.g., the EU AI Act, CSRD, and the U.S. SEC climate disclosure rule), is catalyzing a shift in corporate governance toward stakeholder accountability. Using a curated corpus of seven open-access regulatory and policy texts, we apply a triangulated approach, corpus linguistics (AntConc) and semantic network analysis (InfraNodus), to map how disclosure, risk, assurance, and stakeholder terms structure the discourse. Robustness checks across three stopword specifications (Spec A/B/C) and phrase-level evidence (N-grams/KWIC) corroborate the centrality of disclosure/report/assurance and the conditional peripherality of transparency/accountability. We propose the AI-Policy-Governance Nexus, a conceptual model explaining how regulatory pressure and AI integration reconfigure governance practices beyond compliance. The findings inform strategy, policy design, and future empirical work on AI-enabled ESG systems.